Achievements

Research and translational highlights

The achievements below represent the collective research and translational foundations brought to AIHB by its members; not all outputs were produced under the Centre.

• Prospective and real-world ophthalmic AI: Members have contributed to prospective smartphone screening for childhood visual impairment, multinational AI-based measurement of proptosis in thyroid eye disease, and cross-dataset retinal anomaly detection. Together, these studies span screening, quantitative assessment, and fundus intelligence;
• Digital pathology: Member work includes direct prediction of breast cancer histochemical scores, automated PD-L1 tumour-proportion scoring, nuclei and gland analysis, virtual staining, cross-stain segmentation, and research using public pathology benchmarks such as MIDOG 2022 (MItosis DOmain Generalization Challenge) and DigestPath. The breast cancer H-score work directly addressed the subjectivity and workload of manual pathology scoring;
• Cardiovascular intelligence across modalities: Member foundations span future-event prediction from invasive coronary angiography, left-ventricular segmentation and flow quantification from 4D-flow MRI, prompt-enabled cine-CMR (cardiovascular magnetic resonance) foundation-model adaptation, and cardiac CT wall-motion analysis. The cine-CMR work was evaluated on external, multicentre, and multi-vendor data, with results showing strong agreement between automatically and manually derived clinical parameters;
• Fetal ultrasound and global health technology: Research includes automatic localisation of standard fetal ultrasound clips, congenital heart defect detection, and tools intended to reduce acquisition and interpretation barriers in settings with limited specialist capacity.
• Multi-omics and epitranscriptomics: Members have contributed to deep-learning methods for integrated prediction of RNA modifications, , single-cell epitranscriptomic trajectory inference, and pan-cancer analysis of dysregulated RNA modifications. These capabilities provide a research foundation for connecting clinical phenotypes to biological mechanisms;
• Multimodal mental health and affective intelligence: Member research spans clinical speech screening, multimodal depression assessment, empathetic dialogue, multimodal sentiment analysis, speech emotion recognition, pain detection, and privacy-aware health agents. A 2026 depression study used a large language model (LLM)-powered interview platform and compared the multimodal results with assessments by trained clinicians;
• Embodied emotional support: XJTLU participated in the joint development of An’an, a biomimetic affective-AI panda robot integrating multimodal perception and empathetic interaction. The product was named a CES Innovation Awards 2026 Honoree in the Artificial Intelligence category. In the project, the XJTLU team was principally responsible for AI, intelligent perception, robotics, and human-computer interaction, while the company was responsible for engineering, hardware integration, and manufacturing;
• Intelligent rehabilitation: Member work includes a lower-limb rehabilitation carpet, a smart medical splint, VR-assisted stroke rehabilitation, assistive walking devices, and rehabilitation sensing. Two projects developed by student teams under the supervision of Centre members—A Customisable Lower Limb Rehabilitation Carpet for Children and Smart Medical Splints—won the Grand Prize and Second Prize, respectively, at the Sixth Jiangsu Province Biomedical Engineering Innovation Design Competition
• Medical visualisation, robotics, and biofabrication: The member base includes industrial experience in CT visualisation and surgery navigation, portable head-CT motion correction, mixed-reality microsurgery simulation, robotic needle biopsy, patient-specific 3D reconstruction, intelligent bioscaffold fabrication, and 3D bioprinting;
• Trustworthy algorithms and knowledge systems: Members have developed causal methods for reducing spurious radiological correlations, explainable time-series models, biomedical retrieval with knowledge graphs, continual learning, formal compliance reasoning, and evidence-aware RAG (retrieval-augmented generation) and agent systems; and
• Cross-disciplinary translation: Member projects range from chromosome and cytology analysis to emotional robots, clinical knowledge engines, rehabilitation devices, biological scaffolds, and hospital-facing educational tools. The chromosome-AI work illustrates a translation pathway from interdisciplinary algorithm development to industrial application.

Representative and major papers

Multimodal and multi-omics intelligence for major diseases

• Chen W. et al., “Early Detection of Visual Impairment in Young Children Using a Smartphone-based Deep Learning System,” Nature Medicine, 2023; Kang Dang, co-author.
• Lei C. et al., “ProptoView: AI-based digital exophthalmometry using multi-view facial images in a multinational validation study,” Journal of Translational Medicine, 2026; Kang Dang, co-author.
• Niu J. et al., “HolistAno: Retinal anomaly detection with holistic feature modeling,” Expert Systems with Applications, 2026; Kang Dang, co-first author.
• Lei C. et al., “AI-assisted Facial Analysis in Healthcare: From Disease Detection to Comprehensive Management,” Patterns, 2025; Kang Dang, co-first author.
• Liu J. et al., “An End-to-End Deep Learning Histochemical Scoring System for Breast Cancer TMA,” IEEE Transactions on Medical Imaging, 2019; Jingxin Liu, first author.
• Liu J. et al., “Automated Tumor Proportion Score Analysis for PD-L1 (22C3) Expression in Lung Squamous Cell Carcinoma,” Scientific Reports, 2021; Jingxin Liu, first author.
• Aubreville M. et al., “Mitosis Domain Generalization in Histopathology Images: The MIDOG Challenge,” Medical Image Analysis, 2023; Jingxin Liu, contributor.
• Da Q. et al., “DigestPath: A Benchmark Dataset with Challenge Review for the Pathological Detection and Segmentation of Digestive-System,” Medical Image Analysis, 2022; Jingxin Liu, contributor.
• Sun X. et al., “Future Cardiovascular Events Prediction from Invasive Coronary Angiography: A Graph Representation Learning Perspective,” Medical Image Analysis, 2026; Xiaowu Sun, first author.
• Sun X. et al., “Deep Learning Based Automated Left Ventricle Segmentation and Flow Quantification in 4D Flow Cardiac MRI,” Journal of Cardiovascular Magnetic Resonance, 2024; Xiaowu Sun, first author.
• Chen Z. et al., “Cine Cardiac Magnetic Resonance Segmentation Using Temporal-Spatial Adaptation of Prompt-enabled Segment-Anything-Model: A Feasibility Study,” Journal of Cardiovascular Magnetic Resonance, 2025; Zhennong Chen, co-first author.
• Mishra D. et al., “MCAT: Visual Query-Based Localization of Standard Anatomical Clips in Fetal Ultrasound Videos Using Multi-Tier Class-Aware Token Transformer,” AAAI, 2025; Netzahualcoyotl Hernandez-Cruz, co-author.
• Meng R. et al., “MAST-Pro: Dynamic Mixture-of-Experts for Adaptive Segmentation of Pan-Tumors with Knowledge-Driven Prompts,” MICCAI, 2025; Sifan Song, co-first author.
• Song S. et al., “DualStreamFoveaNet: A Dual Stream Fusion Architecture with Anatomical Awareness for Robust Fovea Localization,” IEEE Journal of Biomedical and Health Informatics, 2024; Sifan Song, first author.
• Song Z. et al., “Attention-based Multi-label Neural Networks for Integrated Prediction and Interpretation of Twelve Widely Occurring RNA Modifications,” Nature Communications, 2021; Jionglong Su, co-author and project-conception contributor.
• Huang D. et al., “Geographic Encoding of Transcripts Enabled High-Accuracy and Isoform-Aware Deep Learning of RNA Methylation,” Nucleic Acids Research, 2022; Jionglong Su, co-author.
• Wang H. et al., “Statistical Modeling of Single-Cell Epitranscriptomics Enabled Trajectory and Regulatory Inference of RNA Methylation,” Cell Genomics, 2024; Jionglong Su, co-author.

Trustworthy biomedical intelligence and smart medical technologies

• Zeng X. et al., “Beyond Shortcuts: Mitigating Spurious Correlations in Radiological Diagnosis with Causal Intervention,” Knowledge-Based Systems, 2026; Haiyang Zhang, corresponding author.
• Chen Q. and Zhang H. et al., “PU-Bench: A Unified Benchmark for Rigorous and Reproducible Positive–Unlabelled Learning,” ICLR, 2026; Haiyang Zhang, co-first and corresponding author.
• Wang Y. et al., “Biomedical Information Retrieval with Positive–Unlabelled Learning and Knowledge Graphs,” ACM Transactions on Intelligent Systems and Technology, 2024; Wei Wang, co-author.
• Wang C. et al., “DAED: Dynamic Additive Effect Decomposition for Interpretable Time-Series Forecasting,” DASFAA, 2026; Chaoqun Wang, first and corresponding author.
• Wang C. et al., “DeLELSTM: Decomposition-based Linear Explainable LSTM to Capture Instantaneous and Long-term Effects in Time Series,” IJCAI, 2023; Chaoqun Wang, first author.
• Lam H.-P. and Hashmi M., “Enabling Reasoning with LegalRuleML,” Theory and Practice of Logic Programming, 2019; an important foundation for formal, auditable and rule-based decision systems.
• Xiang N. and collaborators, mixed-reality microsurgery simulation, The Visual Computer, 2023; part of the centre’s medical-simulation foundation.
• Wang F. et al., “Transformer Based Tissue Classification in Robotic Needle Biopsy,” IEEE SMC, 2024; Fanxin Wang, first author.
• Wang F. et al., “Needle Biopsy and Fiber-Optic Compatible Robotic Insertion Platform,” IEEE EMBC, 2025; Fanxin Wang, first author.

Affective computing, healthy ageing and intelligent care

• Chen Y. et al., “Depression Detection via Multimodal Analysis Using a Large Language Model-powered Interview Platform,” Engineering Applications of Artificial Intelligence, 2026; Yangbin Chen, first author.
• He X. et al., “AL-HCL: Active Learning and Hierarchical Contrastive Learning for Multimodal Sentiment Analysis with Fusion Guidance,” IEEE Transactions on Affective Computing, 2026; Yushan Pan, corresponding author.
• He X. et al., “An Active Learning-based Alternative Reinforcement Contextual Information Fusion Model for Multimodal Sentiment Analysis,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2025; Yushan Pan, corresponding author.
• Ma W. and Jin N., “Multi-scale Local–Global Fusion Network with Temporal Attention for Speech Emotion Recognition,” Neural Computing and Applications, 2026; Nanlin Jin, corresponding author.
• Zhang K. et al., customisable lower-limb rehabilitation carpet for children with cerebral palsy, ACM Interactive Surfaces and Spaces, 2024; supervised by Jie Sun and Qinglei Bu.
• Bu Q. and collaborators, DeepSeek-guided virtual-reality stroke rehabilitation, AIxVR, 2026.
• Bu Q. and collaborators, digital traditional-Chinese-medicine splint and intelligent rehabilitation monitoring, IEEE TrustCom, 2024.

Patents, software copyrights and technology transfer

Publication control: Patent totals below are dossier counts and may include applications, granted patents, shared inventions and records from previous employment. Counts and ownership must be reconciled with the XJTLU Technology Transfer Office before publication.

• Kang Dang: 11 Chinese invention patents, 7 granted, covering medical-image and healthcare-AI technologies.
• Jionglong Su: Member-submitted CV reports one granted and 14 pending patents during the previous five years, including work related to cytology, chromosome analysis and medical imaging.
• Jie Sun: Broad patent portfolio in rehabilitation, intelligent medical devices, biofabrication and related systems, with three technology-licensing records reported in the submitted dossier.
• Qinglei Bu: Three granted rehabilitation-related patents and six applications under substantive examination reported in the submitted material.
• Yangbin Chen: Three granted Chinese patents: CN116312970B for intelligent interactive psychological assessment; CN115995116B for computer-vision-based depression assessment; and CN115934909B for empathetic response generation.
• Fanxin Wang: CN202210469156.9, granted, for point-cloud motion compensation; CN201510571758.5, granted, for high-response drive technology; and CN202510268551.4, pending, for UAV control.
• Zhijie Xu: Four granted Chinese patents listed in the supplied dossier, including CN112215185B, CN114723951B, CN113362457B and CN111582122B. XJTLU’s platform profile states that he holds 11 international patents across his career.
• Pengjing Xu: US10925481B2 and CN115730593B, together with three registered software copyrights for visual-function measurement systems.
• Hong Seng Gan: Four software copyrights reported for medical-image and multimodal-AI systems, including registrations made in 2021, 2022 and 2026.
• Sifan Song: Three medical-imaging patent records related to cervical-cancer detection and chromosome-image segmentation.
• Zhennong Chen: US20250152119A1 for automated wall-motion abnormality evaluation and US20230289972A1 for deep-learning cardiac segmentation and motion visualisation.
• Mian Zhou: Granted or filed inventions in generative data augmentation, multi-object tracking, visual description, bird’s-eye-view perception and occupancy modelling.
• Jianjia Wang: Patent application 202510344327.9 concerning graph-convolutional drug repositioning, in addition to granted inventions related to complex-network analysis.
• Fanxin Wang, Jie Sun, Qinglei Bu, Pengjing Xu, Lingxiao Zhao, Nan Xiang and related members: Collectively provide an intellectual-property base extending from algorithms into medical devices, rehabilitation, visualisation, robotics and manufacturing.

Awards, talent programmes and newsworthy achievements

• CES Innovation Awards 2026: An’an was named an Honoree in the Artificial Intelligence category. The recognition provides a strong international translation story for embodied emotional intelligence.
• Sixth Jiangsu Province Biomedical Engineering Innovation Design Competition: The lower-limb rehabilitation carpet received the Grand Prize and the smart medical splint received the Second Prize.
• ACM ISS 2024: The rehabilitation-carpet work received a Best Poster award according to the submitted member records.
• Kang Dang: Area Chair for MICCAI 2025 and 2026; leadership experience in industrial medical-AI algorithm teams.
• Xiaowu Sun: The 4D-flow cardiac-MRI paper was selected as a finalist for the JCMR Gerald Pohost Award.
• Zhennong Chen: RSNA Trainee Research Prize 2024 and SPIE Medical Imaging Best Student Paper Finalist 2022.
• Sifan Song: ISBI Best Paper Award Finalist for Bilateral-ViT and second place in the MICCAI GAMMA Challenge.
• Wei Wang and collaborating teams: First place in a FinNLP/EMNLP 2025 shared task, third prize in the NAACL CLPsych 2025 shared task, COLING 2025 Best Short Paper and second prize in a CCL 2024 evaluation task.
• Nan Xiang and Yushan Pan: Suzhou Association for Artificial Intelligence Natural Science First Prize reported in the submitted records.
• Zhijie Xu: XJTLU’s official platform profile lists recognition under a national high-level talent programme in 2025, SIP Science and Education Leading Talent in 2024, the Sichuan Thousand Talents Plan and a Sichuan Science and Technology Progress Award.
• Jingxin Liu: Gusu Youth Innovation Leading Talent, Jiangsu Dual-Innovation Doctor, Shanghai Baoshan May First Labour Medal and Shanghai Baoshan Artisan recognition.
• Pengjing Xu and Jianjia Wang: Shanghai Pujiang Talent programme recipients.
• Yushan Pan, Chaoqun Wang, Hejia Qiu and Fanxin Wang: Jiangsu Dual-Innovation Doctor recognition reported in their submitted records.
• Lingxiao Zhao: Chinese Academy of Sciences Hundred Talents background and substantial medical-product development experience at Philips Healthcare and Philips Research.